A lot of the PI drawbacks / justifications for splitting sound like they would be eliminated with the introduction of a cut operator similar to Prolog. My limit

A lot of the PI drawbacks / justifications for splitting sound like they would be eliminated with the introduction of a cut operator similar to Prolog. My limited experience has not seen any evidence of cut operators being used in language processing; are there downsides that only become evident with experience?